Intravoxel incoherent motion diffusion-weighted imaging for the assessment of renal injury in cirrhotic patients

Abstract

Background: Renal dysfunction is a common complication in patients with cirrhosis, and early detection is crucial for timely intervention and treatment. Intravoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI) serves as a non-invasive imaging technique that provides valuable insights into tissue perfusion and diffusion changes, demonstrating significant superiority in assessing renal injury. This prospective study aimed to evaluate early renal injury in patients with cirrhosis using IVIM DWI and to explore the correlation of IVIM parameters with the severity of liver cirrhosis based on the Child-Pugh classification.

Methods: Sixty-four cirrhotic patients and 30 healthy subjects underwent IVIM on a 3.0-T magnetic resonance imaging (MRI). Diffusion coefficient (ADC_slow), pseudo-diffusion coefficient (ADC_fast), and perfusion fraction (f) were derived from the bi-exponential model. In the control group, IVIM-derived parameters for both renal cortex and medulla were compared between right and left kidneys. Subsequently, IVIM-derived renal cortical and medullary parameters were compared between the cirrhotic and control groups. Additionally, within the cirrhotic group, IVIM-derived renal parameters were correlated with the Child-Pugh classification.

Results: In the control group, bilateral renal IVIM parameters (ADC_slow, ADC_fast, and f) showed no significant differences between left and right kidneys (all P>0.05), justifying the use of averaged values for subsequent analysis. Comparative assessment between cirrhotic patients and controls revealed significantly lower renal cortical ADC_fast (P=0.033) and f values (P=0.049) in the cirrhotic group, while ADC_slow did not differ significantly (P=0.846). In the renal medulla, cirrhotic patients exhibited reduced ADC_fast (P=0.043) compared with controls, with no differences in ADC_slow or f (P=0.638 and 0.173, respectively). Notably, renal cortical ADC_fast demonstrated a significant inverse correlation with Child-Pugh classification (R=-0.406, P=0.001), being higher in Child-Pugh A vs. B (P=0.032) and A vs. C (P=0.019) patients, whereas ADC_slow and f showed no correlation with hepatic function grade (P=0.817 and 0.191). Similarly, renal medullary ADC_fast correlated inversely with Child-Pugh classification (R=-0.251, P=0.045), with higher values in class A vs. B (P=0.017), but no differences between class A vs. C (P=0.052) or B vs. C (P=0.448). ADC_slow and f values in both cortical and medullary regions remained non-significant across Child-Pugh classification (all P>0.05).

Conclusions: IVIM DWI non-invasively assesses early renal injury in cirrhotic patients, with reduced renal perfusion correlating with liver cirrhosis severity.

Keywords: Cirrhosis; diffusion-weighted imaging (DWI); intravoxel incoherent motion (IVIM); kidneys; magnetic resonance imaging (MRI).

Figure 1

ROIs were delineated in the IVIM images with a b-value of 10 s/mm². (A) ROI 1 and ROI 2 correspond to the freely hand-drawn ROIs (each with an area of 5–9 cm²) on the right and left kidneys, respectively, at the level of the renal hilum, both completely covering the renal cortex. (B) ROI 1–3 delineate the renal medulla region of the right kidney, while ROI 4–6 delineate the renal medulla region of the left kidney, with each set containing three manually placed elliptical ROIs (each with an area of 0.2–0.4 cm²). IVIM, intravoxel incoherent motion; ROI, region of interest.

Figure 2

Study flow diagram. DWI, diffusion-weighted imaging; IVIM, intravoxel incoherent motion; MRI, magnetic resonance imaging.

Figure 3

The box-and-whisker plots, including individual data points, displaying the comparison of ADC_slow (A,D), ADC_fast (B,E), and f (C,F) values between the renal cortex and medulla in both the right and left kidneys within the control group. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction.

Figure 4

Axial IVIM imaging and parameter maps from a 50-year-old healthy female subject. (A) IVIM image with a b-value of 10 s/mm²; (B) IVIM-derived ADC_slow map; (C) ADC_fast map; (D) f map. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction; IVIM, intravoxel incoherent motion; RI, rainbow invert.

Figure 5

The box-and-whisker plots, incorporating individual data points, illustrating the comparison of (A-C) cortical and (D-F) medullary ADC_slow, ADC_fast, and f values between the control group and the cirrhotic group, respectively. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction.

Figure 6

Correlations between IVIM-derived renal (A-C) cortical and (D-F) medullary parameters and Child-Pugh classification. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction; IVIM, intravoxel incoherent motion.

Figure 7

The box-and-whisker plot with individual data points of the (A,D) ADC_slow, (B,E) ADC_fast, and (C,F) f values among different severity of cirrhotic according to the Child-Pugh classification. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction.

Figure 8

The axial IVIM images with a b-value of 10 s/mm² (A,E,I) and IVIM-derived ADC_slow (B,F,J), ADC_fast (C,G,K), and f (D,H,L) maps from three representative cirrhotic patients. (A-D) A 53-year-old female with liver cirrhosis classified as class A according to the Child-Pugh classification. (E-H) A 52-year-old man with liver cirrhosis classified as class B. (I-L) A 54-year-old man with liver cirrhosis classified as class C. ADC_fast, pseudo-diffusion coefficient; ADC_slow, diffusion coefficient; f, perfusion fraction; IVIM, intravoxel incoherent motion; RI, rainbow invert.